The present thesis addressed several hypotheses and experimental questions regarding psychophysiological and subjective correlates of affective stimulus elaboration in TBI patients with mainly orbitofrontal lesions. Furthermore, the study aimed at investigating whether the extent of prefrontal brain damage and the presence of lesions in ventromedial prefrontal areas or in temporal lobe regions were associated with specific impairments in processing emotionally relevant contents.
This last section summarizes main findings of the present investigation by emphasizing their relevance for the initial hypotheses and experimental aims. The contribution of these findings to the understanding of the neuroanatomical basis of emotional evaluation is discussed, together with the implications they have for the assumptions of current theories of emotion.
Elaboration of emotional stimuli in TBI patients differed in many aspects from that of healthy persons pointing to an impaired discrimination between emotionally salient and neutral contents in these patients. Patients’ ERPs showed a reduced cortical positivity in response to all pictorial stimuli, and furthermore, were not consistently modulated as a function of emotional arousal. The general attenuation of the P3 component and of the subsequent late positive potential suggests an impaired ability to direct and maintain sufficient attentional resources on picture processing, irrespectively of the emotional relevance of stimuli. Another facet of this impairment may be the severe memory deficit for previously presented pictures. Beside this unspecific impairment in picture elaboration, findings also confirmed the reduced ERP discrimination between arousing and neutral stimuli in brain injured patients that characterized the entire late picture processing stage (650ms – 4s). In particular, the missing enhancement of late negativity over occipital areas in response to highly arousing stimuli suggests that, in patients, emotional slides were not associated with a more intense level of visual elaboration, when compared to less relevant pictures.
Patients’ evoked potentials at frontal sites beyond P3 did not differentiate between unpleasant and neutral pictures, whereas pleasant slides elicited an enhanced cortical positivity. This specific abnormality in evaluating unpleasant stimuli was further reflected by subjective data with patients’
SAM arousal ratings for unpleasant pictures being noticeably reduced compared to scorings of healthy subjects. Considering that the lesions of the present head injured patients were mainly localized in orbitofrontal cortical areas, the particular impairment for the elaboration of negative emotional contents found here is not completely unexpected. Both functional imaging studies (George et al., 1995; Northoff et al., 2000) and a single case study of a frontal injured patient
(Angrilli et al., 1999) argue for a stronger involvement of orbitofrontal brain regions in processing of unpleasant stimuli, compared to pleasant ones. Taken together with the present findings, this evidence indicates that elaboration of unpleasant compared to pleasant emotions underlies, at least to some extent, different neuroanatomical substrates.
Evidence from the present subgroup comparisons provided further support for the assumption that prefrontal areas play a key role in the processing of unpleasant stimuli. In fact, both the presence of ventromedial prefrontal lesions and the extent of frontal brain damage were associated with a more prominent reduction of late event related positivity during viewing of unpleasant pictures. Instead, the P3 component and subsequent slow wave potentials were not modulated as a function of stimulus arousal in patients with temporal lobe lesions. Together, these finding suggest that whereas temporal areas seem to have an equally important top down influence on visual processing of both unpleasant and pleasant stimuli, the relevance of orbitofrontal areas for the emotion modulating network appears to be mainly related to negative stimulation. This conclusion would be in line with evidence from functional imaging studies that attribute similar roles to temporal (Kuniecki et al., 2003) and orbitofrontal cortical areas (Lane et al., 1997; Northoff et al., 2000) in visual affective processing. With respect to the influence of the size of frontal lesions, initially, it was assumed that very extensive lesions would be associated with a rather unspecific impairment of cognitive capacities and subjective and psychophysiological indices of emotional elaboration The comparison between patients with large and small frontal lesions has confirmed this hypothesis by revealing that subjects with more extensive brain damage showed, beside their specific deficit in processing unpleasant stimuli, signs of a more general decline of cognitive as well as physiological measures.
In fact, patients with large frontal lesions, compared to the less damaged subgroup, were characterized by greater impairments of incidental memory and cognitive fluency and by generally reduced skin conductance responsiveness to pictorial stimuli.
In general, the present findings on skin conductance reactions were not able to confirm the experimental hypothesis assuming reduced SC responses to emotionally salient stimuli in patients with frontal brain lesions. Our contradictory evidence has been extensively discussed in preceding chapters where the suggestion has already been introduced in which the intact electrodermal performance in patients might be related to the fact that the anterior cingulate cortex, known as a key structure in modulating SC responses to emotional cues, was not damaged in the present clinical sample. However, considering the fact that both evoked potentials and subjective arousal ratings revealed abnormal responses to emotional stimuli in brain lesioned patients, one might also
argue that skin conductance responsiveness may not be such a reliable indicator of emotional impairment as previously assumed.
Damasio and colleagues have repeatedly emphasized the lack of SC responses to affective cues as a key deficit in patients with frontal lesions. In fact, Damasio considered his “unequivocal” results (Damasio, 1994, p. 209) as the main confirming evidence for his “somatic marker theory”. The specific SCR impairment in frontal patients was interpreted as the incapacity of those patients, when being presented with an emotional stimulus, to retrieve past emotional experience which should lead to a physiological activation appropriate to the emotional cue. In Damasio’s words, they
“could not produce a somatic state or, in the very least, a somatic state of which they could be aware” (Damasio, 1994, p.211). In the present thesis, the skin conductance findings along with evidence from ERP and subjective data argues against the assumption of the complete failure in frontally damaged patients to adequately respond to affective stimuli. Even though in patients with prefrontal lesions, both parameters revealed prominent impairments in elaborating emotionally relevant pictures, it is important to note that neither the whole patient group nor any of the lesion subgroups showed an entirely abolished distinction between affective and neutral stimuli. In particular, the ratings of self-experienced arousal reveal that, although emotional stimuli, especially unpleasant ones, were perceived as less exciting, frontally lesioned patients were still able to experience an emotional state. Thus, the term “somatic marker deficit” should be used with caution and, in particular, should not be generally applied to all frontal lobe patients. Evidence from the present study suggests that distinct areas within the frontal lobe, for instance the VMPFC, might be related to specific impairments of emotional elaboration. In this regard it would be interesting for future studies to investigate affective stimulus processing in frontally damaged patients with and without damage of the anterior cingulate cortex.
An alternative explanation for patients’ particular impairment of emotional elaboration could be provided by referring to the motivational organization of emotion as proposed by Lang (Lang, 1998; Lang et al., 1997). In this regard, both physiological and subjective data would argue for a specific impairment in frontally damaged patients to adequately direct motivated attention towards emotionally relevant stimuli. Very early ERP responses that differ between arousing and neutral pictures and the capacity to produce adequate SCRs indicate that patients are principally able to differentiate between emotional and neutral contents. However, the consistent impairments during later stages of stimulus elaboration together with the reduction in self-perceived arousal suggest that emotional pictures are considered by patients as motivationally less relevant. As a consequence, less attentional resources are allocated to these stimuli which leads to a less deep and efficient cognitive
elaboration process. Our data further show that this attention-related deficit in patients with prefrontal lesions might be more pronounced with respect to aversive emotional motivation compared to appetitive motivation.
Besides revealing that traumatic brain lesions lead to significant deficits of affective stimulus evaluation, the present findings provide important insights with respect to temporal aspects of emotional elaboration by allowing conclusions about which brain structures are involved at what specific processing stage. Our ERP results strongly support the assumption of a network of various neuroanatomical structures associated with the elaboration of emotionally relevant stimuli. In agreement with previous functional imaging studies, our evidence suggests that this network recruits medial temporal areas as the superior and middle temporal gyrus (Iidaka et al., 2001) and parts of the prefrontal cortex, namely ventromedial prefrontal areas (Northoff et al., 2000) and the orbitofrontal cortex (Keightley, Winocur, Graham, Mayberg et al., 2002). Since amygdala lesions could not be detected in the brain injured patients, it was not possible to investigate the specific effect of such lesions on affective elaboration and draw reliable conclusions about the role of limbic structures in emotion modulation. However, the finding that very rapid discrimination between arousing and calm pictures was not affected in the TBI patient group might be an indication that at this early processing stage, the amygdala may play an especially important role whereas orbitofrontal areas are not involved. This assumption is supported by studies arguing for a modulatory effect of the amygdala on early affective perceptual processing (Anderson & Phelps, 2001) and the amygdala’s implication in non-conscious monitoring of emotional stimuli (Whalen et al., 1998). Future clinical investigations should address this issue by studying rapid EEG responses to emotional stimuli in patients with well defined amygdala lesions.
In addition to the amygdala, evidence from the lesion group comparisons suggests the ventromedial prefrontal cortex as a further brain structure associated with rapid picture processing. In fact, the 160-220 ms ERP pattern of patients with VMPFC lesions was found to reflect less efficient visual encoding of highly arousing slides compared to patients without lesions in this brain structure. This leads to the suggestion that ventromedial prefrontal areas may be able to modulate even relatively early aspects of perceptual processing of emotional stimuli via top-down influences on the visual cortex. Support for this notion comes from single-unit recordings in human VMPFC in response to presentation of affective facial expression (Kawasaki et al., 2001).
Conscious cognitive processing of affective stimuli as indexed by the P3 and subsequent slow potentials was found to be noticeably affected by prefrontal (ventromedial and orbitofrontal) and
temporal cortical lesions. The middle and superior temporal gyri that receive bilateral projections from the amygdala, and prefrontal regions which are reciprocally connected with visual association areas have already been reported to be part of a network that modulates evaluation of facial expression and emotional pictures (Iidaka et al., 2001; Kuniecki et al., 2003). The present evidence would therefore fit the idea that once a stimulus is initially identified as emotionally salient, prefrontal contribute to stimulus relevance by relating information about the external visual cue to interoceptive information. If the stimulus is categorized as motivationally relevant, the prefrontal cortex may trigger the allocation of attentional resources towards this stimulus and finally, through his reciprocal connections with visual association areas in the temporal lobe (see Figure 6.1.), prime the visual cortex for further perceptual processing. This top-down influence of the prefrontal cortex is also described by Lang (1998) who explains that the “increased activity…in perceptual processing areas for emotional … stimuli implies reentrant processing from sites more anterior in the brain”. Our data furthermore indicate, that the modulatory effect of frontal brain areas might depend on stimulus valence; the specific impairment at later processing stages in differentiating between unpleasant and neutral pictures that was found for the whole patient group and also for the subgroup with VMPFC lesions would argue for a particular involvement of prefrontal areas in evaluating unpleasant emotional stimuli.
My conclusion follows with a graphical illustration (Figure 6.1.) adapted from a recent review by Adolphs (2002) who has summarized various literature on perceptual processing of emotion.
Although his conclusions were mainly based on studies on recognition of emotional facial expressions, they offer several similarities to the evidence found in the present thesis with respect to both temporal as well as neuroanatomical aspects of visual affective processing. Figure 6.1. shows, on the left, Adolph’s schematic illustration of the initial stages of emotional stimulus elaboration and, on the right, the respective supporting evidence from the present thesis.
Time course of Evidence from the decisive role at this early stage
=> Amygdala as a possible key structure for pre-attentive, automatic emotional processing
=> VMPFC involved in modulation of rapid visual affective
discrimination (160-220 ms)
=> Top-down influence of prefrontal areas (orbitofrontal, VMPFC) on attention-
modulated visual processing of particularly unpleasant stimuli
=> Involvement of temporal areas (middle and superior temporal gyri) in the modulation of visual processing of both types of emotional stimuli (pleasant and unpleasant)
adapted from Adolphs, 2002
(a) (b)
Fig. 6.1. Time course and neuroanatomical correlates of visual affective processing
(a) Processing of emotional facial expression as a function of time, and brain structures involved at various time points; model proposed by Adolphs (2002)
(A=Amygdala, FFA: fusiform face area; INS: insula; O: orbitofrontal cortex; SC:
superior colliculus; SCx: striate cortex; SS: somatosensory cortex; STG: superior temporal gyrus; T: Thalamus)
(b) Evidence from the present investigation argues for similar temporal and neuro-anatomical mechanisms in the processing of visual affective stimuli other than faces
Outlook
This thesis aimed, for the first time, at investigating deficits of emotional processing in TBI patients by systematically studying physiological and subjective responses to standardized emotional visual stimuli in a representative sample of head injured patients with mainly frontal lesions. Results support a model of an impaired discrimination between neutral and emotionally relevant stimuli in these patients and point to a specific role of prefrontal brain areas in affective picture processing.
Methodological limitations of the present study should provide recommendations for further research on brain injured patients:
The clinical sample was selected to be homogeneous with respect to rehabilitation status and brain lesion characteristics in that all TBI patients had predominantly orbitofrontal lesions and showed no substantial impairment of those basic cognitive functions (orientation, perception, and language comprehension) necessary to understand the experimental procedure. This homogeneity was advantageous when comparing the whole patient sample with a control group. It implied, however, some problems regarding the generation of subgroups. It was not possible to divide patients into small groups with completely distinguished lesions; instead, there was still a substantial lesion overlap between subgroups that were compared with each other. The absence of differences between two clinical samples, for instance, on neuropsychological measures and skin conductance responses, might, be related to their shared neuroanatomical and basic cognitive characteristics. To draw a reliable conclusion about the specific effect of a given brain structure on affective stimulus processing, future studies should compare representative samples of different head injured groups characterized by well defined lesions. Ideally, patients with circumscribed lesions of the amygdala and the anterior cingulate gyrus should also be included, as the consequences of damage to these brain areas that are assumed to play an important role for emotional elaboration, could not be investigated in the present patient group. A further issue which could not be addressed concerns the impact of lesion lateralization on affective stimulus evaluation. As our patients had all suffered bilateral, rather symmetrical brain damage, a further subdivision into groups with left and right hemispheric lesions was not possible. Future research on patients with unilateral lesions would be particularly useful to gain important insights in hemispheric specialization for emotional processing.
The ERP methodology employed for the present experimental paradigm provides an excellent temporal resolution and is particularly suited for studying cortical activity changes by relating them to different stages during picture viewing. To obtain more information on spatial characteristics of emotional processing, further EEG-studies should use high-density electrode arrays, together with a distributed source analysis procedure to reliably estimate the cortical sources of changes in the
scalp-recorded ERP. Another recommendable approach would be the use of whole head Magnetoencephalography (MEG) where localization of activity sources is possible even with greater confidence.
Recent research has demonstrated that high-speed presentation of up to 5Hz already leads to rapid detection and perceptual processing of emotionally salient stimuli (Junghöfer et al., 2001). Such a rapid projection of standardized pictures could be advantageous for future studies with brain injured patients as well as other clinical populations. First, many pictures can be presented in a very brief period, thus abbreviating the experimental session, and second, the investigation of emotional processing may be possible without having to fully activate potentially distressing affective responses in patients.
When discussing future research directions, implications for TBI rehabilitation programs should be taken into account. The present study showed that neuropsychological deficits are not necessarily severely impaired after brain damage, whereas the elaboration of emotional stimuli is characterized by consistent alterations compared to healthy persons. If further clinical investigations will be able to come to similar conclusions and provide additional knowledge about the quality and the neuroanatomical etiology of emotional alterations in TBI patients, important insights into effective treatment approaches may be gained. Instead of focusing – as it is the case today - mainly on the rehabilitation of cognitive skills, optimal treatment of brain injured individuals rather should consist of an integrated approach to the wide spectrum of possible neuropsychological, emotional, and behavioral dysfunctions. In addition, relatives of the patients should be included in treatment programs, as the daily family life or the relationship with a partner are often particularly affected by emotional alterations that follow a TBI. Enhanced knowledge about consequences that specific lesions have for emotional behavior might increase the awareness of patients and their family about changed ways of reacting to certain stimuli and situations in brain injured individuals. Furthermore, on the basis of this knowledge, individualized rehabilitation programs might be designed which enhance the probability that a TBI patient will be able to regain his cognitive, social and professional functionality.
7. REFERENCES
Adams, J.H., Graham, D.I., Murray, L.S. & Scott, G. (1982). Diffuse axonal injury due to nonmissile head injury in humans: An analysis of 45 cases. Annals of Neurology, 12, 557-563.
Adams, J.H., Graham., D.I., Scott, G., Parker, L.S., Doyle, D. (1980). Brain damage in fatal non-missile head injury. J Clin Pathol, 33, 1132-1145.
Adolphs, R. (2002). Neural systems for recognizing emotion. Current Opinion in Neurobiology, 12, 1-9.
Adolphs, R., Baron-Cohen, S., Tranel, D. (2002). Impaired Recognition of social emotions following amygdala damage. Journal of Cognitive Neuroscience, 14(8), 1264-1274.
Adolphs, R., Damasio, H.., Tranel, D. & Damasio A.R. (1996). Cortical systems for the recognition of emotion in facial expression. The Journal of Neuroscience, 16(23), 7678-7687.
Adolphs, R. & Tranel, D. (2003). Amygdala damage impairs emotion recognition from scenes only when they contain facial expression. Neuropsychologia, 41, 1281-1289.
Adolphs, R., Tranel, D., Damasio, H., Damasio, A.R. (1994). Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala. Nature, 372, 669-672.
Adolphs, R., Tranel, D., Hamann, S., Young, A.W., Calder, A.J., Phelps, E.A., Anderson, A., Lee,G.P., Damasio, A.R. (1999). Recognition of facial emotion in nine individuals with bilateral amygdale damage. Neuropsychologia, 37, 1111-1117.
Aggleton, J.P. (1992). The functional effects of amygdala lesions in humans: a comparison with
Aggleton, J.P. (1992). The functional effects of amygdala lesions in humans: a comparison with